Professional Documents
Culture Documents
Perceptron Convergance Theorem
Perceptron Convergance Theorem
Theorem
Perceptron: Limitations
Proof:
• suppose x C1 output = 1 and x C2 output = -1.
• For simplicity assume w(1) = 0, = 1.
• Suppose perceptron incorrectly classifies x(1) … x(n) … C1.
Then wT(k) x(k) 0.
Error correction rule:
w(2) = w(1) + x(1)
w(3) = w(2) + x(2)
w(n+1) = x(1)+ …+ x(n)
w(n+1) = w(n) + x(n).
Convergence theorem (proof)
• Cauchy-Schwarz inequality:
||w0||2 ||w(n+1)||2 [w0T w(n+1)]2
n2 2
||w(n+1)||2 ||w0|| 2 (A)
Convergence theorem (proof)
• Now we consider another route:
• w(k+1) = w(k) + x(k) ||
w(k+1)||2 = || w(k)||2 + ||x(k)||2 + 2 w T(k)x(k)
euclidean norm
0 because x(k) is misclassified
||w(k+1)||2 ||w(k)||2 + ||x(k)||2 k=1,..,n
=0
||w(2)||2 ||w(1)||2 + ||x(1)||2
||w(3)||2 ||w(2)||2 + ||x(2)||2
n
x(k )
2
||w(n+1)||2
k 1
convergence theorem (proof)
• Let = max ||x(n)||2 x(n) C1
• ||w(n+1)||2 n
• For (B) sufficiently large values of k:
• (B) becomes in conflict with (A). Then n cannot be greater than nmax such
that (A) and (B) are both satisfied with the equality sign.
n 2max 2 || w0 || 2
2
nmax nmax 2
β
|| w0 ||
• Perceptron convergence algorithm terminates in at most iterations.
nmax=
||w0||2
2
Adaline: Adaptive Linear Element
x1
w1
y
x2 w2
wm
m
xm
y x j (n)w j (n )
j 0
Adaline: Adaptive Linear Element
• Adaline: uses a linear neuron model and the Least-Mean-Square (LMS)
learning algorithm
The idea: try to minimize the square error, which is a function of the
weights
E ( w(n)) 12 e 2 ( n )
m
e( n ) d ( n ) x j (n)w j (n )
j 0
(gradient of E ( w )) E , , E
w1 wm
• make a small step in that direction
• Approximation of gradient(E)
E ( w(n)) e(n)
e(n)
w(n) w(n)
e(n)[ x(n) T ]